Merge branch 'linus' into x86/urgent, to refresh the tree

Signed-off-by: Ingo Molnar <mingo@kernel.org>

16 files changed, 385 insertions, 140 deletions

diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c
index 7b76eb67a9b3..c343a54bed39 100644
--- a/arch/x86/kernel/cpu/amd.c
+++ b/arch/x86/kernel/cpu/amd.c
@@ -565,14 +565,17 @@ static void early_init_amd(struct cpuinfo_x86 *c)
 	 * can safely set X86_FEATURE_EXTD_APICID unconditionally for families
 	 * after 16h.
 	 */
-	if (cpu_has_apic && c->x86 > 0x16) {
-		set_cpu_cap(c, X86_FEATURE_EXTD_APICID);
-	} else if (cpu_has_apic && c->x86 >= 0xf) {
-		/* check CPU config space for extended APIC ID */
-		unsigned int val;
-		val = read_pci_config(0, 24, 0, 0x68);
-		if ((val & ((1 << 17) | (1 << 18))) == ((1 << 17) | (1 << 18)))
+	if (boot_cpu_has(X86_FEATURE_APIC)) {
+		if (c->x86 > 0x16)
 			set_cpu_cap(c, X86_FEATURE_EXTD_APICID);
+		else if (c->x86 >= 0xf) {
+			/* check CPU config space for extended APIC ID */
+			unsigned int val;
+
+			val = read_pci_config(0, 24, 0, 0x68);
+			if ((val >> 17 & 0x3) == 0x3)
+				set_cpu_cap(c, X86_FEATURE_EXTD_APICID);
+		}
 	}
 #endif
 
@@ -628,6 +631,7 @@ static void init_amd_k8(struct cpuinfo_x86 *c)
 	 */
 	msr_set_bit(MSR_K7_HWCR, 6);
 #endif
+	set_cpu_bug(c, X86_BUG_SWAPGS_FENCE);
 }
 
 static void init_amd_gh(struct cpuinfo_x86 *c)
@@ -746,7 +750,7 @@ static void init_amd(struct cpuinfo_x86 *c)
 	if (c->x86 >= 0xf)
 		set_cpu_cap(c, X86_FEATURE_K8);
 
-	if (cpu_has_xmm2) {
+	if (cpu_has(c, X86_FEATURE_XMM2)) {
 		/* MFENCE stops RDTSC speculation */
 		set_cpu_cap(c, X86_FEATURE_MFENCE_RDTSC);
 	}
diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
index 62ff5255ae16..0fe6953f421c 100644
--- a/arch/x86/kernel/cpu/common.c
+++ b/arch/x86/kernel/cpu/common.c
@@ -437,7 +437,7 @@ void load_percpu_segment(int cpu)
 #ifdef CONFIG_X86_32
 	loadsegment(fs, __KERNEL_PERCPU);
 #else
-	loadsegment(gs, 0);
+	__loadsegment_simple(gs, 0);
 	wrmsrl(MSR_GS_BASE, (unsigned long)per_cpu(irq_stack_union.gs_base, cpu));
 #endif
 	load_stack_canary_segment();
@@ -724,6 +724,13 @@ void get_cpu_cap(struct cpuinfo_x86 *c)
 		}
 	}
 
+	if (c->extended_cpuid_level >= 0x80000007) {
+		cpuid(0x80000007, &eax, &ebx, &ecx, &edx);
+
+		c->x86_capability[CPUID_8000_0007_EBX] = ebx;
+		c->x86_power = edx;
+	}
+
 	if (c->extended_cpuid_level >= 0x80000008) {
 		cpuid(0x80000008, &eax, &ebx, &ecx, &edx);
 
@@ -736,9 +743,6 @@ void get_cpu_cap(struct cpuinfo_x86 *c)
 		c->x86_phys_bits = 36;
 #endif
 
-	if (c->extended_cpuid_level >= 0x80000007)
-		c->x86_power = cpuid_edx(0x80000007);
-
 	if (c->extended_cpuid_level >= 0x8000000a)
 		c->x86_capability[CPUID_8000_000A_EDX] = cpuid_edx(0x8000000a);
 
@@ -869,30 +873,34 @@ static void detect_nopl(struct cpuinfo_x86 *c)
 #else
 	set_cpu_cap(c, X86_FEATURE_NOPL);
 #endif
+}
 
+static void detect_null_seg_behavior(struct cpuinfo_x86 *c)
+{
+#ifdef CONFIG_X86_64
 	/*
-	 * ESPFIX is a strange bug.  All real CPUs have it.  Paravirt
-	 * systems that run Linux at CPL > 0 may or may not have the
-	 * issue, but, even if they have the issue, there's absolutely
-	 * nothing we can do about it because we can't use the real IRET
-	 * instruction.
+	 * Empirically, writing zero to a segment selector on AMD does
+	 * not clear the base, whereas writing zero to a segment
+	 * selector on Intel does clear the base.  Intel's behavior
+	 * allows slightly faster context switches in the common case
+	 * where GS is unused by the prev and next threads.
 	 *
-	 * NB: For the time being, only 32-bit kernels support
-	 * X86_BUG_ESPFIX as such.  64-bit kernels directly choose
-	 * whether to apply espfix using paravirt hooks.  If any
-	 * non-paravirt system ever shows up that does *not* have the
-	 * ESPFIX issue, we can change this.
+	 * Since neither vendor documents this anywhere that I can see,
+	 * detect it directly instead of hardcoding the choice by
+	 * vendor.
+	 *
+	 * I've designated AMD's behavior as the "bug" because it's
+	 * counterintuitive and less friendly.
 	 */
-#ifdef CONFIG_X86_32
-#ifdef CONFIG_PARAVIRT
-	do {
-		extern void native_iret(void);
-		if (pv_cpu_ops.iret == native_iret)
-			set_cpu_bug(c, X86_BUG_ESPFIX);
-	} while (0);
-#else
-	set_cpu_bug(c, X86_BUG_ESPFIX);
-#endif
+
+	unsigned long old_base, tmp;
+	rdmsrl(MSR_FS_BASE, old_base);
+	wrmsrl(MSR_FS_BASE, 1);
+	loadsegment(fs, 0);
+	rdmsrl(MSR_FS_BASE, tmp);
+	if (tmp != 0)
+		set_cpu_bug(c, X86_BUG_NULL_SEG);
+	wrmsrl(MSR_FS_BASE, old_base);
 #endif
 }
 
@@ -928,6 +936,33 @@ static void generic_identify(struct cpuinfo_x86 *c)
 	get_model_name(c); /* Default name */
 
 	detect_nopl(c);
+
+	detect_null_seg_behavior(c);
+
+	/*
+	 * ESPFIX is a strange bug.  All real CPUs have it.  Paravirt
+	 * systems that run Linux at CPL > 0 may or may not have the
+	 * issue, but, even if they have the issue, there's absolutely
+	 * nothing we can do about it because we can't use the real IRET
+	 * instruction.
+	 *
+	 * NB: For the time being, only 32-bit kernels support
+	 * X86_BUG_ESPFIX as such.  64-bit kernels directly choose
+	 * whether to apply espfix using paravirt hooks.  If any
+	 * non-paravirt system ever shows up that does *not* have the
+	 * ESPFIX issue, we can change this.
+	 */
+#ifdef CONFIG_X86_32
+# ifdef CONFIG_PARAVIRT
+	do {
+		extern void native_iret(void);
+		if (pv_cpu_ops.iret == native_iret)
+			set_cpu_bug(c, X86_BUG_ESPFIX);
+	} while (0);
+# else
+	set_cpu_bug(c, X86_BUG_ESPFIX);
+# endif
+#endif
 }
 
 static void x86_init_cache_qos(struct cpuinfo_x86 *c)
@@ -1083,12 +1118,12 @@ void enable_sep_cpu(void)
 	struct tss_struct *tss;
 	int cpu;
 
+	if (!boot_cpu_has(X86_FEATURE_SEP))
+		return;
+
 	cpu = get_cpu();
 	tss = &per_cpu(cpu_tss, cpu);
 
-	if (!boot_cpu_has(X86_FEATURE_SEP))
-		goto out;
-
 	/*
 	 * We cache MSR_IA32_SYSENTER_CS's value in the TSS's ss1 field --
 	 * see the big comment in struct x86_hw_tss's definition.
@@ -1103,7 +1138,6 @@ void enable_sep_cpu(void)
 
 	wrmsr(MSR_IA32_SYSENTER_EIP, (unsigned long)entry_SYSENTER_32, 0);
 
-out:
 	put_cpu();
 }
 #endif
@@ -1535,7 +1569,7 @@ void cpu_init(void)
 	pr_info("Initializing CPU#%d\n", cpu);
 
 	if (cpu_feature_enabled(X86_FEATURE_VME) ||
-	    cpu_has_tsc ||
+	    boot_cpu_has(X86_FEATURE_TSC) ||
 	    boot_cpu_has(X86_FEATURE_DE))
 		cr4_clear_bits(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE);
 
diff --git a/arch/x86/kernel/cpu/cyrix.c b/arch/x86/kernel/cpu/cyrix.c
index 6adef9cac23e..bd9dcd6b712d 100644
--- a/arch/x86/kernel/cpu/cyrix.c
+++ b/arch/x86/kernel/cpu/cyrix.c
@@ -333,7 +333,7 @@ static void init_cyrix(struct cpuinfo_x86 *c)
 		switch (dir0_lsn) {
 		case 0xd:  /* either a 486SLC or DLC w/o DEVID */
 			dir0_msn = 0;
-			p = Cx486_name[(cpu_has_fpu ? 1 : 0)];
+			p = Cx486_name[!!boot_cpu_has(X86_FEATURE_FPU)];
 			break;
 
 		case 0xe:  /* a 486S A step */
diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c
index b47df99dc5d2..6e2ffbebbcdb 100644
--- a/arch/x86/kernel/cpu/intel.c
+++ b/arch/x86/kernel/cpu/intel.c
@@ -187,9 +187,9 @@ static void early_init_intel(struct cpuinfo_x86 *c)
 	 *  the TLB when any changes are made to any of the page table entries.
 	 *  The operating system must reload CR3 to cause the TLB to be flushed"
 	 *
-	 * As a result cpu_has_pge() in arch/x86/include/asm/tlbflush.h should
-	 * be false so that __flush_tlb_all() causes CR3 insted of CR4.PGE
-	 * to be modified
+	 * As a result, boot_cpu_has(X86_FEATURE_PGE) in arch/x86/include/asm/tlbflush.h
+	 * should be false so that __flush_tlb_all() causes CR3 insted of CR4.PGE
+	 * to be modified.
 	 */
 	if (c->x86 == 5 && c->x86_model == 9) {
 		pr_info("Disabling PGE capability bit\n");
@@ -270,7 +270,7 @@ static void intel_workarounds(struct cpuinfo_x86 *c)
 	 * The Quark is also family 5, but does not have the same bug.
 	 */
 	clear_cpu_bug(c, X86_BUG_F00F);
-	if (!paravirt_enabled() && c->x86 == 5 && c->x86_model < 9) {
+	if (c->x86 == 5 && c->x86_model < 9) {
 		static int f00f_workaround_enabled;
 
 		set_cpu_bug(c, X86_BUG_F00F);
@@ -318,7 +318,7 @@ static void intel_workarounds(struct cpuinfo_x86 *c)
 	 * integrated APIC (see 11AP erratum in "Pentium Processor
 	 * Specification Update").
 	 */
-	if (cpu_has_apic && (c->x86<<8 | c->x86_model<<4) == 0x520 &&
+	if (boot_cpu_has(X86_FEATURE_APIC) && (c->x86<<8 | c->x86_model<<4) == 0x520 &&
 	    (c->x86_mask < 0x6 || c->x86_mask == 0xb))
 		set_cpu_bug(c, X86_BUG_11AP);
 
@@ -493,7 +493,7 @@ static void init_intel(struct cpuinfo_x86 *c)
 			set_cpu_cap(c, X86_FEATURE_ARCH_PERFMON);
 	}
 
-	if (cpu_has_xmm2)
+	if (cpu_has(c, X86_FEATURE_XMM2))
 		set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC);
 
 	if (boot_cpu_has(X86_FEATURE_DS)) {
@@ -505,7 +505,7 @@ static void init_intel(struct cpuinfo_x86 *c)
 			set_cpu_cap(c, X86_FEATURE_PEBS);
 	}
 
-	if (c->x86 == 6 && cpu_has_clflush &&
+	if (c->x86 == 6 && boot_cpu_has(X86_FEATURE_CLFLUSH) &&
 	    (c->x86_model == 29 || c->x86_model == 46 || c->x86_model == 47))
 		set_cpu_bug(c, X86_BUG_CLFLUSH_MONITOR);
 
diff --git a/arch/x86/kernel/cpu/mcheck/mce-genpool.c b/arch/x86/kernel/cpu/mcheck/mce-genpool.c
index 2658e2af74ec..93d824ec3120 100644
--- a/arch/x86/kernel/cpu/mcheck/mce-genpool.c
+++ b/arch/x86/kernel/cpu/mcheck/mce-genpool.c
@@ -26,6 +26,52 @@ static struct gen_pool *mce_evt_pool;
 static LLIST_HEAD(mce_event_llist);
 static char gen_pool_buf[MCE_POOLSZ];
 
+/*
+ * Compare the record "t" with each of the records on list "l" to see if
+ * an equivalent one is present in the list.
+ */
+static bool is_duplicate_mce_record(struct mce_evt_llist *t, struct mce_evt_llist *l)
+{
+	struct mce_evt_llist *node;
+	struct mce *m1, *m2;
+
+	m1 = &t->mce;
+
+	llist_for_each_entry(node, &l->llnode, llnode) {
+		m2 = &node->mce;
+
+		if (!mce_cmp(m1, m2))
+			return true;
+	}
+	return false;
+}
+
+/*
+ * The system has panicked - we'd like to peruse the list of MCE records
+ * that have been queued, but not seen by anyone yet.  The list is in
+ * reverse time order, so we need to reverse it. While doing that we can
+ * also drop duplicate records (these were logged because some banks are
+ * shared between cores or by all threads on a socket).
+ */
+struct llist_node *mce_gen_pool_prepare_records(void)
+{
+	struct llist_node *head;
+	LLIST_HEAD(new_head);
+	struct mce_evt_llist *node, *t;
+
+	head = llist_del_all(&mce_event_llist);
+	if (!head)
+		return NULL;
+
+	/* squeeze out duplicates while reversing order */
+	llist_for_each_entry_safe(node, t, head, llnode) {
+		if (!is_duplicate_mce_record(node, t))
+			llist_add(&node->llnode, &new_head);
+	}
+
+	return new_head.first;
+}
+
 void mce_gen_pool_process(void)
 {
 	struct llist_node *head;
diff --git a/arch/x86/kernel/cpu/mcheck/mce-internal.h b/arch/x86/kernel/cpu/mcheck/mce-internal.h
index 547720efd923..cd74a3f00aea 100644
--- a/arch/x86/kernel/cpu/mcheck/mce-internal.h
+++ b/arch/x86/kernel/cpu/mcheck/mce-internal.h
@@ -35,6 +35,7 @@ void mce_gen_pool_process(void);
 bool mce_gen_pool_empty(void);
 int mce_gen_pool_add(struct mce *mce);
 int mce_gen_pool_init(void);
+struct llist_node *mce_gen_pool_prepare_records(void);
 
 extern int (*mce_severity)(struct mce *a, int tolerant, char **msg, bool is_excp);
 struct dentry *mce_get_debugfs_dir(void);
@@ -81,3 +82,17 @@ static inline int apei_clear_mce(u64 record_id)
 #endif
 
 void mce_inject_log(struct mce *m);
+
+/*
+ * We consider records to be equivalent if bank+status+addr+misc all match.
+ * This is only used when the system is going down because of a fatal error
+ * to avoid cluttering the console log with essentially repeated information.
+ * In normal processing all errors seen are logged.
+ */
+static inline bool mce_cmp(struct mce *m1, struct mce *m2)
+{
+	return m1->bank != m2->bank ||
+		m1->status != m2->status ||
+		m1->addr != m2->addr ||
+		m1->misc != m2->misc;
+}
diff --git a/arch/x86/kernel/cpu/mcheck/mce-severity.c b/arch/x86/kernel/cpu/mcheck/mce-severity.c
index 5119766d9889..631356c8cca4 100644
--- a/arch/x86/kernel/cpu/mcheck/mce-severity.c
+++ b/arch/x86/kernel/cpu/mcheck/mce-severity.c
@@ -204,6 +204,33 @@ static int error_context(struct mce *m)
 	return IN_KERNEL;
 }
 
+static int mce_severity_amd_smca(struct mce *m, int err_ctx)
+{
+	u32 addr = MSR_AMD64_SMCA_MCx_CONFIG(m->bank);
+	u32 low, high;
+
+	/*
+	 * We need to look at the following bits:
+	 * - "succor" bit (data poisoning support), and
+	 * - TCC bit (Task Context Corrupt)
+	 * in MCi_STATUS to determine error severity.
+	 */
+	if (!mce_flags.succor)
+		return MCE_PANIC_SEVERITY;
+
+	if (rdmsr_safe(addr, &low, &high))
+		return MCE_PANIC_SEVERITY;
+
+	/* TCC (Task context corrupt). If set and if IN_KERNEL, panic. */
+	if ((low & MCI_CONFIG_MCAX) &&
+	    (m->status & MCI_STATUS_TCC) &&
+	    (err_ctx == IN_KERNEL))
+		return MCE_PANIC_SEVERITY;
+
+	 /* ...otherwise invoke hwpoison handler. */
+	return MCE_AR_SEVERITY;
+}
+
 /*
  * See AMD Error Scope Hierarchy table in a newer BKDG. For example
  * 49125_15h_Models_30h-3Fh_BKDG.pdf, section "RAS Features"
@@ -225,6 +252,9 @@ static int mce_severity_amd(struct mce *m, int tolerant, char **msg, bool is_exc
 		 * to at least kill process to prolong system operation.
 		 */
 		if (mce_flags.overflow_recov) {
+			if (mce_flags.smca)
+				return mce_severity_amd_smca(m, ctx);
+
 			/* software can try to contain */
 			if (!(m->mcgstatus & MCG_STATUS_RIPV) && (ctx == IN_KERNEL))
 				return MCE_PANIC_SEVERITY;
diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c
index f0c921b03e42..92e5e37d97bf 100644
--- a/arch/x86/kernel/cpu/mcheck/mce.c
+++ b/arch/x86/kernel/cpu/mcheck/mce.c
@@ -161,7 +161,6 @@ void mce_log(struct mce *mce)
 	if (!mce_gen_pool_add(mce))
 		irq_work_queue(&mce_irq_work);
 
-	mce->finished = 0;
 	wmb();
 	for (;;) {
 		entry = mce_log_get_idx_check(mcelog.next);
@@ -194,7 +193,6 @@ void mce_log(struct mce *mce)
 	mcelog.entry[entry].finished = 1;
 	wmb();
 
-	mce->finished = 1;
 	set_bit(0, &mce_need_notify);
 }
 
@@ -224,6 +222,53 @@ void mce_unregister_decode_chain(struct notifier_block *nb)
 }
 EXPORT_SYMBOL_GPL(mce_unregister_decode_chain);
 
+static inline u32 ctl_reg(int bank)
+{
+	return MSR_IA32_MCx_CTL(bank);
+}
+
+static inline u32 status_reg(int bank)
+{
+	return MSR_IA32_MCx_STATUS(bank);
+}
+
+static inline u32 addr_reg(int bank)
+{
+	return MSR_IA32_MCx_ADDR(bank);
+}
+
+static inline u32 misc_reg(int bank)
+{
+	return MSR_IA32_MCx_MISC(bank);
+}
+
+static inline u32 smca_ctl_reg(int bank)
+{
+	return MSR_AMD64_SMCA_MCx_CTL(bank);
+}
+
+static inline u32 smca_status_reg(int bank)
+{
+	return MSR_AMD64_SMCA_MCx_STATUS(bank);
+}
+
+static inline u32 smca_addr_reg(int bank)
+{
+	return MSR_AMD64_SMCA_MCx_ADDR(bank);
+}
+
+static inline u32 smca_misc_reg(int bank)
+{
+	return MSR_AMD64_SMCA_MCx_MISC(bank);
+}
+
+struct mca_msr_regs msr_ops = {
+	.ctl	= ctl_reg,
+	.status	= status_reg,
+	.addr	= addr_reg,
+	.misc	= misc_reg
+};
+
 static void print_mce(struct mce *m)
 {
 	int ret = 0;
@@ -290,7 +335,9 @@ static void wait_for_panic(void)
 
 static void mce_panic(const char *msg, struct mce *final, char *exp)
 {
-	int i, apei_err = 0;
+	int apei_err = 0;
+	struct llist_node *pending;
+	struct mce_evt_llist *l;
 
 	if (!fake_panic) {
 		/*
@@ -307,11 +354,10 @@ static void mce_panic(const char *msg, struct mce *final, char *exp)
 		if (atomic_inc_return(&mce_fake_panicked) > 1)
 			return;
 	}
+	pending = mce_gen_pool_prepare_records();
 	/* First print corrected ones that are still unlogged */
-	for (i = 0; i < MCE_LOG_LEN; i++) {
-		struct mce *m = &mcelog.entry[i];
-		if (!(m->status & MCI_STATUS_VAL))
-			continue;
+	llist_for_each_entry(l, pending, llnode) {
+		struct mce *m = &l->mce;
 		if (!(m->status & MCI_STATUS_UC)) {
 			print_mce(m);
 			if (!apei_err)
@@ -319,13 +365,11 @@ static void mce_panic(const char *msg, struct mce *final, char *exp)
 		}
 	}
 	/* Now print uncorrected but with the final one last */
-	for (i = 0; i < MCE_LOG_LEN; i++) {
-		struct mce *m = &mcelog.entry[i];
-		if (!(m->status & MCI_STATUS_VAL))
-			continue;
+	llist_for_each_entry(l, pending, llnode) {
+		struct mce *m = &l->mce;
 		if (!(m->status & MCI_STATUS_UC))
 			continue;
-		if (!final || memcmp(m, final, sizeof(struct mce))) {
+		if (!final || mce_cmp(m, final)) {
 			print_mce(m);
 			if (!apei_err)
 				apei_err = apei_write_mce(m);
@@ -356,11 +400,11 @@ static int msr_to_offset(u32 msr)
 
 	if (msr == mca_cfg.rip_msr)
 		return offsetof(struct mce, ip);
-	if (msr == MSR_IA32_MCx_STATUS(bank))
+	if (msr == msr_ops.status(bank))
 		return offsetof(struct mce, status);
-	if (msr == MSR_IA32_MCx_ADDR(bank))
+	if (msr == msr_ops.addr(bank))
 		return offsetof(struct mce, addr);
-	if (msr == MSR_IA32_MCx_MISC(bank))
+	if (msr == msr_ops.misc(bank))
 		return offsetof(struct mce, misc);
 	if (msr == MSR_IA32_MCG_STATUS)
 		return offsetof(struct mce, mcgstatus);
@@ -523,9 +567,9 @@ static struct notifier_block mce_srao_nb = {
 static void mce_read_aux(struct mce *m, int i)
 {
 	if (m->status & MCI_STATUS_MISCV)
-		m->misc = mce_rdmsrl(MSR_IA32_MCx_MISC(i));
+		m->misc = mce_rdmsrl(msr_ops.misc(i));
 	if (m->status & MCI_STATUS_ADDRV) {
-		m->addr = mce_rdmsrl(MSR_IA32_MCx_ADDR(i));
+		m->addr = mce_rdmsrl(msr_ops.addr(i));
 
 		/*
 		 * Mask the reported address by the reported granularity.
@@ -607,7 +651,7 @@ bool machine_check_poll(enum mcp_flags flags, mce_banks_t *b)
 		m.tsc = 0;
 
 		barrier();
-		m.status = mce_rdmsrl(MSR_IA32_MCx_STATUS(i));
+		m.status = mce_rdmsrl(msr_ops.status(i));
 		if (!(m.status & MCI_STATUS_VAL))
 			continue;
 
@@ -654,7 +698,7 @@ bool machine_check_poll(enum mcp_flags flags, mce_banks_t *b)
 		/*
 		 * Clear state for this bank.
 		 */
-		mce_wrmsrl(MSR_IA32_MCx_STATUS(i), 0);
+		mce_wrmsrl(msr_ops.status(i), 0);
 	}
 
 	/*
@@ -679,7 +723,7 @@ static int mce_no_way_out(struct mce *m, char **msg, unsigned long *validp,
 	char *tmp;
 
 	for (i = 0; i < mca_cfg.banks; i++) {
-		m->status = mce_rdmsrl(MSR_IA32_MCx_STATUS(i));
+		m->status = mce_rdmsrl(msr_ops.status(i));
 		if (m->status & MCI_STATUS_VAL) {
 			__set_bit(i, validp);
 			if (quirk_no_way_out)
@@ -830,9 +874,9 @@ static int mce_start(int *no_way_out)
 
 	atomic_add(*no_way_out, &global_nwo);
 	/*
-	 * global_nwo should be updated before mce_callin
+	 * Rely on the implied barrier below, such that global_nwo
+	 * is updated before mce_callin.
 	 */
-	smp_wmb();
 	order = atomic_inc_return(&mce_callin);
 
 	/*
@@ -957,7 +1001,7 @@ static void mce_clear_state(unsigned long *toclear)
 
 	for (i = 0; i < mca_cfg.banks; i++) {
 		if (test_bit(i, toclear))
-			mce_wrmsrl(MSR_IA32_MCx_STATUS(i), 0);
+			mce_wrmsrl(msr_ops.status(i), 0);
 	}
 }
 
@@ -994,11 +1038,12 @@ void do_machine_check(struct pt_regs *regs, long error_code)
 	int i;
 	int worst = 0;
 	int severity;
+
 	/*
 	 * Establish sequential order between the CPUs entering the machine
 	 * check handler.
 	 */
-	int order;
+	int order = -1;
 	/*
 	 * If no_way_out gets set, there is no safe way to recover from this
 	 * MCE.  If mca_cfg.tolerant is cranked up, we'll try anyway.
@@ -1012,7 +1057,12 @@ void do_machine_check(struct pt_regs *regs, long error_code)
 	DECLARE_BITMAP(toclear, MAX_NR_BANKS);
 	DECLARE_BITMAP(valid_banks, MAX_NR_BANKS);
 	char *msg = "Unknown";
-	int lmce = 0;
+
+	/*
+	 * MCEs are always local on AMD. Same is determined by MCG_STATUS_LMCES
+	 * on Intel.
+	 */
+	int lmce = 1;
 
 	/* If this CPU is offline, just bail out. */
 	if (cpu_is_offline(smp_processor_id())) {
@@ -1051,19 +1101,20 @@ void do_machine_check(struct pt_regs *regs, long error_code)
 		kill_it = 1;
 
 	/*
-	 * Check if this MCE is signaled to only this logical processor
+	 * Check if this MCE is signaled to only this logical processor,
+	 * on Intel only.
 	 */
-	if (m.mcgstatus & MCG_STATUS_LMCES)
-		lmce = 1;
-	else {
-		/*
-		 * Go through all the banks in exclusion of the other CPUs.
-		 * This way we don't report duplicated events on shared banks
-		 * because the first one to see it will clear it.
-		 * If this is a Local MCE, then no need to perform rendezvous.
-		 */
+	if (m.cpuvendor == X86_VENDOR_INTEL)
+		lmce = m.mcgstatus & MCG_STATUS_LMCES;
+
+	/*
+	 * Go through all banks in exclusion of the other CPUs. This way we
+	 * don't report duplicated events on shared banks because the first one
+	 * to see it will clear it. If this is a Local MCE, then no need to
+	 * perform rendezvous.
+	 */
+	if (!lmce)
 		order = mce_start(&no_way_out);
-	}
 
 	for (i = 0; i < cfg->banks; i++) {
 		__clear_bit(i, toclear);
@@ -1076,7 +1127,7 @@ void do_machine_check(struct pt_regs *regs, long error_code)
 		m.addr = 0;
 		m.bank = i;
 
-		m.status = mce_rdmsrl(MSR_IA32_MCx_STATUS(i));
+		m.status = mce_rdmsrl(msr_ops.status(i));
 		if ((m.status & MCI_STATUS_VAL) == 0)
 			continue;
 
@@ -1420,7 +1471,6 @@ static void __mcheck_cpu_init_generic(void)
 	enum mcp_flags m_fl = 0;
 	mce_banks_t all_banks;
 	u64 cap;
-	int i;
 
 	if (!mca_cfg.bootlog)
 		m_fl = MCP_DONTLOG;
@@ -1436,14 +1486,19 @@ static void __mcheck_cpu_init_generic(void)
 	rdmsrl(MSR_IA32_MCG_CAP, cap);
 	if (cap & MCG_CTL_P)
 		wrmsr(MSR_IA32_MCG_CTL, 0xffffffff, 0xffffffff);
+}
+
+static void __mcheck_cpu_init_clear_banks(void)
+{
+	int i;
 
 	for (i = 0; i < mca_cfg.banks; i++) {
 		struct mce_bank *b = &mce_banks[i];
 
 		if (!b->init)
 			continue;
-		wrmsrl(MSR_IA32_MCx_CTL(i), b->ctl);
-		wrmsrl(MSR_IA32_MCx_STATUS(i), 0);
+		wrmsrl(msr_ops.ctl(i), b->ctl);
+		wrmsrl(msr_ops.status(i), 0);
 	}
 }
 
@@ -1495,7 +1550,7 @@ static int __mcheck_cpu_apply_quirks(struct cpuinfo_x86 *c)
 			 */
 			clear_bit(10, (unsigned long *)&mce_banks[4].ctl);
 		}
-		if (c->x86 <= 17 && cfg->bootlog < 0) {
+		if (c->x86 < 17 && cfg->bootlog < 0) {
 			/*
 			 * Lots of broken BIOS around that don't clear them
 			 * by default and leave crap in there. Don't log:
@@ -1628,11 +1683,19 @@ static void __mcheck_cpu_init_vendor(struct cpuinfo_x86 *c)
 		break;
 
 	case X86_VENDOR_AMD: {
-		u32 ebx = cpuid_ebx(0x80000007);
+		mce_flags.overflow_recov = !!cpu_has(c, X86_FEATURE_OVERFLOW_RECOV);
+		mce_flags.succor	 = !!cpu_has(c, X86_FEATURE_SUCCOR);
+		mce_flags.smca		 = !!cpu_has(c, X86_FEATURE_SMCA);
 
-		mce_flags.overflow_recov = !!(ebx & BIT(0));
-		mce_flags.succor	 = !!(ebx & BIT(1));
-		mce_flags.smca		 = !!(ebx & BIT(3));
+		/*
+		 * Install proper ops for Scalable MCA enabled processors
+		 */
+		if (mce_flags.smca) {
+			msr_ops.ctl	= smca_ctl_reg;
+			msr_ops.status	= smca_status_reg;
+			msr_ops.addr	= smca_addr_reg;
+			msr_ops.misc	= smca_misc_reg;
+		}
 		mce_amd_feature_init(c);
 
 		break;
@@ -1717,6 +1780,7 @@ void mcheck_cpu_init(struct cpuinfo_x86 *c)
 
 	__mcheck_cpu_init_generic();
 	__mcheck_cpu_init_vendor(c);
+	__mcheck_cpu_init_clear_banks();
 	__mcheck_cpu_init_timer();
 }
 
@@ -2082,7 +2146,7 @@ static void mce_disable_error_reporting(void)
 		struct mce_bank *b = &mce_banks[i];
 
 		if (b->init)
-			wrmsrl(MSR_IA32_MCx_CTL(i), 0);
+			wrmsrl(msr_ops.ctl(i), 0);
 	}
 	return;
 }
@@ -2121,6 +2185,7 @@ static void mce_syscore_resume(void)
 {
 	__mcheck_cpu_init_generic();
 	__mcheck_cpu_init_vendor(raw_cpu_ptr(&cpu_info));
+	__mcheck_cpu_init_clear_banks();
 }
 
 static struct syscore_ops mce_syscore_ops = {
@@ -2138,6 +2203,7 @@ static void mce_cpu_restart(void *data)
 	if (!mce_available(raw_cpu_ptr(&cpu_info)))
 		return;
 	__mcheck_cpu_init_generic();
+	__mcheck_cpu_init_clear_banks();
 	__mcheck_cpu_init_timer();
 }
 
@@ -2413,7 +2479,7 @@ static void mce_reenable_cpu(void *h)
 		struct mce_bank *b = &mce_banks[i];
 
 		if (b->init)
-			wrmsrl(MSR_IA32_MCx_CTL(i), b->ctl);
+			wrmsrl(msr_ops.ctl(i), b->ctl);
 	}
 }
 
diff --git a/arch/x86/kernel/cpu/mcheck/mce_amd.c b/arch/x86/kernel/cpu/mcheck/mce_amd.c
index 9d656fd436ef..10b0661651e0 100644
--- a/arch/x86/kernel/cpu/mcheck/mce_amd.c
+++ b/arch/x86/kernel/cpu/mcheck/mce_amd.c
@@ -54,14 +54,6 @@
 /* Threshold LVT offset is at MSR0xC0000410[15:12] */
 #define SMCA_THR_LVT_OFF	0xF000
 
-/*
- * OS is required to set the MCAX bit to acknowledge that it is now using the
- * new MSR ranges and new registers under each bank. It also means that the OS
- * will configure deferred errors in the new MCx_CONFIG register. If the bit is
- * not set, uncorrectable errors will cause a system panic.
- */
-#define SMCA_MCAX_EN_OFF	0x1
-
 static const char * const th_names[] = {
 	"load_store",
 	"insn_fetch",
@@ -333,7 +325,7 @@ static u32 get_block_address(u32 current_addr, u32 low, u32 high,
 	/* Fall back to method we used for older processors: */
 	switch (block) {
 	case 0:
-		addr = MSR_IA32_MCx_MISC(bank);
+		addr = msr_ops.misc(bank);
 		break;
 	case 1:
 		offset = ((low & MASK_BLKPTR_LO) >> 21);
@@ -351,6 +343,7 @@ prepare_threshold_block(unsigned int bank, unsigned int block, u32 addr,
 			int offset, u32 misc_high)
 {
 	unsigned int cpu = smp_processor_id();
+	u32 smca_low, smca_high, smca_addr;
 	struct threshold_block b;
 	int new;
 
@@ -369,24 +362,49 @@ prepare_threshold_block(unsigned int bank, unsigned int block, u32 addr,
 
 	b.interrupt_enable = 1;
 
-	if (mce_flags.smca) {
-		u32 smca_low, smca_high;
-		u32 smca_addr = MSR_AMD64_SMCA_MCx_CONFIG(bank);
+	if (!mce_flags.smca) {
+		new = (misc_high & MASK_LVTOFF_HI) >> 20;
+		goto set_offset;
+	}
 
-		if (!rdmsr_safe(smca_addr, &smca_low, &smca_high)) {
-			smca_high |= SMCA_MCAX_EN_OFF;
-			wrmsr(smca_addr, smca_low, smca_high);
-		}
+	smca_addr = MSR_AMD64_SMCA_MCx_CONFIG(bank);
 
-		/* Gather LVT offset for thresholding: */
-		if (rdmsr_safe(MSR_CU_DEF_ERR, &smca_low, &smca_high))
-			goto out;
+	if (!rdmsr_safe(smca_addr, &smca_low, &smca_high)) {
+		/*
+		 * OS is required to set the MCAX bit to acknowledge that it is
+		 * now using the new MSR ranges and new registers under each
+		 * bank. It also means that the OS will configure deferred
+		 * errors in the new MCx_CONFIG register. If the bit is not set,
+		 * uncorrectable errors will cause a system panic.
+		 *
+		 * MCA_CONFIG[MCAX] is bit 32 (0 in the high portion of the MSR.)
+		 */
+		smca_high |= BIT(0);
 
-		new = (smca_low & SMCA_THR_LVT_OFF) >> 12;
-	} else {
-		new = (misc_high & MASK_LVTOFF_HI) >> 20;
+		/*
+		 * SMCA logs Deferred Error information in MCA_DE{STAT,ADDR}
+		 * registers with the option of additionally logging to
+		 * MCA_{STATUS,ADDR} if MCA_CONFIG[LogDeferredInMcaStat] is set.
+		 *
+		 * This bit is usually set by BIOS to retain the old behavior
+		 * for OSes that don't use the new registers. Linux supports the
+		 * new registers so let's disable that additional logging here.
+		 *
+		 * MCA_CONFIG[LogDeferredInMcaStat] is bit 34 (bit 2 in the high
+		 * portion of the MSR).
+		 */
+		smca_high &= ~BIT(2);
+
+		wrmsr(smca_addr, smca_low, smca_high);
 	}
 
+	/* Gather LVT offset for thresholding: */
+	if (rdmsr_safe(MSR_CU_DEF_ERR, &smca_low, &smca_high))
+		goto out;
+
+	new = (smca_low & SMCA_THR_LVT_OFF) >> 12;
+
+set_offset:
 	offset = setup_APIC_mce_threshold(offset, new);
 
 	if ((offset == new) && (mce_threshold_vector != amd_threshold_interrupt))
@@ -430,12 +448,23 @@ void mce_amd_feature_init(struct cpuinfo_x86 *c)
 		deferred_error_interrupt_enable(c);
 }
 
-static void __log_error(unsigned int bank, bool threshold_err, u64 misc)
+static void
+__log_error(unsigned int bank, bool deferred_err, bool threshold_err, u64 misc)
 {
+	u32 msr_status = msr_ops.status(bank);
+	u32 msr_addr = msr_ops.addr(bank);
 	struct mce m;
 	u64 status;
 
-	rdmsrl(MSR_IA32_MCx_STATUS(bank), status);
+	WARN_ON_ONCE(deferred_err && threshold_err);
+
+	if (deferred_err && mce_flags.smca) {
+		msr_status = MSR_AMD64_SMCA_MCx_DESTAT(bank);
+		msr_addr = MSR_AMD64_SMCA_MCx_DEADDR(bank);
+	}
+
+	rdmsrl(msr_status, status);
+
 	if (!(status & MCI_STATUS_VAL))
 		return;
 
@@ -448,10 +477,11 @@ static void __log_error(unsigned int bank, bool threshold_err, u64 misc)
 		m.misc = misc;
 
 	if (m.status & MCI_STATUS_ADDRV)
-		rdmsrl(MSR_IA32_MCx_ADDR(bank), m.addr);
+		rdmsrl(msr_addr, m.addr);
 
 	mce_log(&m);
-	wrmsrl(MSR_IA32_MCx_STATUS(bank), 0);
+
+	wrmsrl(msr_status, 0);
 }
 
 static inline void __smp_deferred_error_interrupt(void)
@@ -479,17 +509,21 @@ asmlinkage __visible void smp_trace_deferred_error_interrupt(void)
 /* APIC interrupt handler for deferred errors */
 static void amd_deferred_error_interrupt(void)
 {
-	u64 status;
 	unsigned int bank;
+	u32 msr_status;
+	u64 status;
 
 	for (bank = 0; bank < mca_cfg.banks; ++bank) {
-		rdmsrl(MSR_IA32_MCx_STATUS(bank), status);
+		msr_status = (mce_flags.smca) ? MSR_AMD64_SMCA_MCx_DESTAT(bank)
+					      : msr_ops.status(bank);
+
+		rdmsrl(msr_status, status);
 
 		if (!(status & MCI_STATUS_VAL) ||
 		    !(status & MCI_STATUS_DEFERRED))
 			continue;
 
-		__log_error(bank, false, 0);
+		__log_error(bank, true, false, 0);
 		break;
 	}
 }
@@ -544,7 +578,7 @@ static void amd_threshold_interrupt(void)
 	return;
 
 log:
-	__log_error(bank, true, ((u64)high << 32) | low);
+	__log_error(bank, false, true, ((u64)high << 32) | low);
 }
 
 /*
diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel.c b/arch/x86/kernel/cpu/mcheck/mce_intel.c
index 1e8bb6c94f14..1defb8ea882c 100644
--- a/arch/x86/kernel/cpu/mcheck/mce_intel.c
+++ b/arch/x86/kernel/cpu/mcheck/mce_intel.c
@@ -84,7 +84,7 @@ static int cmci_supported(int *banks)
 	 */
 	if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL)
 		return 0;
-	if (!cpu_has_apic || lapic_get_maxlvt() < 6)
+	if (!boot_cpu_has(X86_FEATURE_APIC) || lapic_get_maxlvt() < 6)
 		return 0;
 	rdmsrl(MSR_IA32_MCG_CAP, cap);
 	*banks = min_t(unsigned, MAX_NR_BANKS, cap & 0xff);
diff --git a/arch/x86/kernel/cpu/mcheck/therm_throt.c b/arch/x86/kernel/cpu/mcheck/therm_throt.c
index ac780cad3b86..6b9dc4d18ccc 100644
--- a/arch/x86/kernel/cpu/mcheck/therm_throt.c
+++ b/arch/x86/kernel/cpu/mcheck/therm_throt.c
@@ -450,7 +450,7 @@ asmlinkage __visible void smp_trace_thermal_interrupt(struct pt_regs *regs)
 /* Thermal monitoring depends on APIC, ACPI and clock modulation */
 static int intel_thermal_supported(struct cpuinfo_x86 *c)
 {
-	if (!cpu_has_apic)
+	if (!boot_cpu_has(X86_FEATURE_APIC))
 		return 0;
 	if (!cpu_has(c, X86_FEATURE_ACPI) || !cpu_has(c, X86_FEATURE_ACC))
 		return 0;
diff --git a/arch/x86/kernel/cpu/mtrr/cyrix.c b/arch/x86/kernel/cpu/mtrr/cyrix.c
index f8c81ba0b465..b1086f79e57e 100644
--- a/arch/x86/kernel/cpu/mtrr/cyrix.c
+++ b/arch/x86/kernel/cpu/mtrr/cyrix.c
@@ -137,7 +137,7 @@ static void prepare_set(void)
 	u32 cr0;
 
 	/*  Save value of CR4 and clear Page Global Enable (bit 7)  */
-	if (cpu_has_pge) {
+	if (boot_cpu_has(X86_FEATURE_PGE)) {
 		cr4 = __read_cr4();
 		__write_cr4(cr4 & ~X86_CR4_PGE);
 	}
@@ -170,7 +170,7 @@ static void post_set(void)
 	write_cr0(read_cr0() & ~X86_CR0_CD);
 
 	/* Restore value of CR4 */
-	if (cpu_has_pge)
+	if (boot_cpu_has(X86_FEATURE_PGE))
 		__write_cr4(cr4);
 }
 
diff --git a/arch/x86/kernel/cpu/mtrr/generic.c b/arch/x86/kernel/cpu/mtrr/generic.c
index 19f57360dfd2..16e37a2581ac 100644
--- a/arch/x86/kernel/cpu/mtrr/generic.c
+++ b/arch/x86/kernel/cpu/mtrr/generic.c
@@ -444,11 +444,24 @@ static void __init print_mtrr_state(void)
 		pr_debug("TOM2: %016llx aka %lldM\n", mtrr_tom2, mtrr_tom2>>20);
 }
 
+/* PAT setup for BP. We need to go through sync steps here */
+void __init mtrr_bp_pat_init(void)
+{
+	unsigned long flags;
+
+	local_irq_save(flags);
+	prepare_set();
+
+	pat_init();
+
+	post_set();
+	local_irq_restore(flags);
+}
+
 /* Grab all of the MTRR state for this CPU into *state */
 bool __init get_mtrr_state(void)
 {
 	struct mtrr_var_range *vrs;
-	unsigned long flags;
 	unsigned lo, dummy;
 	unsigned int i;
 
@@ -481,15 +494,6 @@ bool __init get_mtrr_state(void)
 
 	mtrr_state_set = 1;
 
-	/* PAT setup for BP. We need to go through sync steps here */
-	local_irq_save(flags);
-	prepare_set();
-
-	pat_init();
-
-	post_set();
-	local_irq_restore(flags);
-
 	return !!(mtrr_state.enabled & MTRR_STATE_MTRR_ENABLED);
 }
 
@@ -741,7 +745,7 @@ static void prepare_set(void) __acquires(set_atomicity_lock)
 	wbinvd();
 
 	/* Save value of CR4 and clear Page Global Enable (bit 7) */
-	if (cpu_has_pge) {
+	if (boot_cpu_has(X86_FEATURE_PGE)) {
 		cr4 = __read_cr4();
 		__write_cr4(cr4 & ~X86_CR4_PGE);
 	}
@@ -771,7 +775,7 @@ static void post_set(void) __releases(set_atomicity_lock)
 	write_cr0(read_cr0() & ~X86_CR0_CD);
 
 	/* Restore value of CR4 */
-	if (cpu_has_pge)
+	if (boot_cpu_has(X86_FEATURE_PGE))
 		__write_cr4(cr4);
 	raw_spin_unlock(&set_atomicity_lock);
 }
diff --git a/arch/x86/kernel/cpu/mtrr/main.c b/arch/x86/kernel/cpu/mtrr/main.c
index 10f8d4796240..7d393ecdeee6 100644
--- a/arch/x86/kernel/cpu/mtrr/main.c
+++ b/arch/x86/kernel/cpu/mtrr/main.c
@@ -752,6 +752,9 @@ void __init mtrr_bp_init(void)
 			/* BIOS may override */
 			__mtrr_enabled = get_mtrr_state();
 
+			if (mtrr_enabled())
+				mtrr_bp_pat_init();
+
 			if (mtrr_cleanup(phys_addr)) {
 				changed_by_mtrr_cleanup = 1;
 				mtrr_if->set_all();
@@ -759,8 +762,16 @@ void __init mtrr_bp_init(void)
 		}
 	}
 
-	if (!mtrr_enabled())
+	if (!mtrr_enabled()) {
 		pr_info("MTRR: Disabled\n");
+
+		/*
+		 * PAT initialization relies on MTRR's rendezvous handler.
+		 * Skip PAT init until the handler can initialize both
+		 * features independently.
+		 */
+		pat_disable("MTRRs disabled, skipping PAT initialization too.");
+	}
 }
 
 void mtrr_ap_init(void)
diff --git a/arch/x86/kernel/cpu/mtrr/mtrr.h b/arch/x86/kernel/cpu/mtrr/mtrr.h
index 951884dcc433..6c7ced07d16d 100644
--- a/arch/x86/kernel/cpu/mtrr/mtrr.h
+++ b/arch/x86/kernel/cpu/mtrr/mtrr.h
@@ -52,6 +52,7 @@ void set_mtrr_prepare_save(struct set_mtrr_context *ctxt);
 void fill_mtrr_var_range(unsigned int index,
 		u32 base_lo, u32 base_hi, u32 mask_lo, u32 mask_hi);
 bool get_mtrr_state(void);
+void mtrr_bp_pat_init(void);
 
 extern void set_mtrr_ops(const struct mtrr_ops *ops);
 
diff --git a/arch/x86/kernel/cpu/vmware.c b/arch/x86/kernel/cpu/vmware.c
index 364e58346897..8cac429b6a1d 100644
--- a/arch/x86/kernel/cpu/vmware.c
+++ b/arch/x86/kernel/cpu/vmware.c
@@ -94,7 +94,7 @@ static void __init vmware_platform_setup(void)
  */
 static uint32_t __init vmware_platform(void)
 {
-	if (cpu_has_hypervisor) {
+	if (boot_cpu_has(X86_FEATURE_HYPERVISOR)) {
 		unsigned int eax;
 		unsigned int hyper_vendor_id[3];